LITHIC REDUCTION STAGES
The production and maintenance of flaked lithic tools has been conceptualized as a series of stages through which the tools pass as their forms are altered by the removal of material, creating flakes and other lithic wastes as byproducts. If discarded tools and lithic wastes can be matched with the stages in the reduction sequence at which they were produced, the spatial organization of tool manufacturing, use, and reworking can be inferred.
Michael B. Collins (1975) proposed a model for the reduction process which has commonly been used locally. Collins distinguished five stages (which he termed "steps"): (I) the acquisition of raw material, (II) core preparation and initial reduction, (III) primary trimming, (IV) secondary trimming and shaping; and (V) tool maintenance and modification. Stages III through V were regarded as optional, because they might be omitted in the production of some types of tools or in some situations.
Other stage analyses have focused on specific end products or specific knapping strategies. Biface reduction was analyzed by Errett Callahan (1979) and Elizabeth J. Skinner (1990) in terms of seven stages: obtaining the blank, initial edging, primary thinning, secondary thinning, shaping to preform, finishing, and reworking/rejuvenation. Diagnostics for the stages were based on the cross-sectional shape of the tool, its width-to-thickness ratio, and its flake scar pattern. Less detailed stage distinctions were proposed for the associated debitage: indicators of relatively early-stage work included cortex, simple dorsal scar morphology, remnants of ventral flake bulbs on dorsal surfaces, and single-faceted platforms.
Martin D. Rosen (1982) carried out an unusually detailed analysis of the debitage assemblages from two sites near Jamacha, SDI-4763 and SDI-5066. He considered up to 38 attributes of the flakes and explicitly linked several of the attributes to Collins' stage scheme. He concluded that SDI-5066 contained residues from more early-stage core preparation work, while SDI-4763 represented later-stage shaping of blanks and preforms for bifacial tools. He suggested as statistical indicators of a predominance of early-stage reduction (a) a higher frequency of flakes with widths greater than their lengths, (b) thicker flakes, (c) a greater mean distance from the point of force application on flakes to the line of maximum width, (d) a greater mean distance from the point of force application to the line of maximum thickness, (e) a higher frequency of flakes with cortical platforms or with fewer platform scars, (f) stronger ventral lipping on flakes, (g) a higher frequency of flakes with dorsal cortex, (h) flakes with fewer dorsal scars, (i) a predominance of larger dorsal scars, and (j) weak patterning of dorsal scar orientation.
D. L. True and Paul D. Bouey (1990) discussed the reduction stages represented by non-quartz debitage at the Gladishill Site, a probable San Dieguito camp near Valley Center. The stage analysis was based informally on flake size and the number of dorsal scars. The conclusion was that "little or no primary tool manufacture took place on the site" (True and Bouey 1990:20).
Brian F. Byrd, John R. Cook, and Carol Serr (1993) discussed the general types of lithic reduction activities that were represented at several sites and loci near Jamacha, although their scheme was not explicitly a stage sequence. They took into account the proportions of primary, secondary, and interior flakes (based on cortex), flake size and fragmentation, and the count ratios of debitage to cores and to flaked tools. These considerations suggested that the sites were used primarily for tool manufacturing and resharpening, not for initial core reduction.
The most extensive application of a reduction stage scheme to local debitage analysis has been in conjunction with a nine-type debitage classification system, developed primarily by E. Jane Rosenthal (Norwood et al. 1982) and Susan M. Hector (1984). The interpretation of debitage types in terms of lithic reduction stages was apparently based on replicative experiments as well as general reasoning. In applications of the nine-type system by different investigators, both the precise definitions of the nine types and their reduction-stage interpretations have varied slightly. Several reservations concerning the system's use in inferring lithic reduction stages may be suggested:
-- The validity of its types as indicators of particular reduction stages has apparently not been tested empirically. This might be done by comparing results from this method with other indications of site function or by testing the internal consistency of reduction stage indications derived from the various types.
-- Similarly, the optimality of the specific attributes used in the typology for the purpose of identifying reduction stages has not been tested empirically. For instance, would qualitatively different attributes, different length or dorsal cortex percentage classes, or different ways of clustering the attributes be more successful in making the desired stage distinctions?
-- It is not clear whether the relationship between a lithic waste type and a reduction stage is intended to be interpreted as absolute or statistical. For instance, would the presence of wastes without platforms but with cortex (Type 8) be taken to establish the presence of Stage II or Stage III reduction, or would a relatively high proportion of such wastes merely be
taken to indicate a probable preponderance of those stages?
-- Given that most of the types have been assigned to more than one reduction stage, and given that such multiple assignments overlap, it is not clear how, statistically, the relative importance of a particular stage (or group of stages) within an assemblage should to be expressed.
-- Almost all of the sizeable lithic waste assemblages that have been reported contain all nine waste types, but specific standards have not yet been suggested for deciding what constitutes a relatively heavy or light emphasis on work at a particular reduction stage.
In addition to the traits mentioned above, several other assemblage attributes may be suggested as relevant to the recognition of lithic reduction stages at specific sites:
-- The presence of suitable unworked raw material in the immediate vicinity of a site may be considered to be a necessary attribute for the recognition of Stage I activity.
-- The presence of test blocks (cores with one or two flakes removed from otherwise-cortical surfaces) may be considered a sufficient attribute for the recognition of Stage I activity.
-- The presence of cores (distinguished from test blocks and abandoned blanks or preforms) may be considered a sufficient attribute for the recognition of Stage II activity.
-- The presence of failed blanks may be considered a sufficient attribute for the recognition of Stage II or Stage III activity.
-- The presence of failed preforms may be considered a sufficient attribute for the recognition of Stage III or Stage IV activity.
-- The presence of broken, substantially-finished tools with no use wear may be considered a sufficient attribute for the recognition of Stage IV activity.
-- The presence of flakes with use wear on the dorsal edges of their platforms may be considered a necessary attribute for the recognition of Stage V activity.
PROSPECTS
Future archaeological investigations and replicative experiments may be able to refine the criteria for identifying the presence/absence and relative preponderance of various lithic reduction stages. Relevant evidence is likely to include the absolute sizes and ratios of various artifact dimensions, the proportion and location of cortex, flaking patterns, and material types. Geographical, chronological, and functional contexts of recovered assemblages are also likely to be significant.
Nine-Type System: Type Numbers and Defining Attributes
| Type |
Platform
Present? |
Widest at
Platform? |
Length >=
2 x Width? |
Dorsal
Cortex? |
Length
>3 cm? |
| 1 |
yes |
no |
yes |
-- |
-- |
| 2 |
yes |
yes |
no |
-- |
-- |
| 3 |
yes |
no |
no |
>90% |
yes |
| 4 |
yes |
no |
no |
30-90% |
yes |
| 5 |
yes |
no |
no |
<30% |
yes |
| 6 |
yes |
no |
no |
absent |
no |
| 7 |
yes |
no |
no |
present |
no |
| 8 |
no |
-- |
-- |
present |
-- |
| 9 |
no |
-- |
-- |
absent |
-- |
Type Numbers and Associated Stages, According to Analysts
|
Hector 1984 |
Rosen 1984 |
Dominici 1985 |
Corum 1986 |
| 1 |
III |
III |
III |
III |
| 2 |
III, IV |
III, IV |
III, IV |
III, IV |
| 3 |
II, III |
II, III |
II, III |
II, III |
| 4 |
II, III |
II, III |
II, III |
II, III |
| 5 |
II |
II, III |
II, III |
II, III |
| 6 |
IV, V |
IV, V |
IV, V |
IV, V |
| 7 |
IV, V |
II, III, IV |
II, III |
II, IV |
| 8 |
II |
II, III |
II, III |
II, III |
| 9 |
IV |
IV, V |
IV, V |
IV, V |